Add raw text option to text conversion (lesmianator will need it).

[librarian.git] / librarian / pdf.py

diff --git a/librarian/pdf.py b/librarian/pdf.py
index 5debb68..5f6f0a2 100644 (file)
--- a/librarian/pdf.py
+++ b/librarian/pdf.py
@@ -21,7 +21,7 @@ from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 from librarian.dcparser import Person
 from librarian.parser import WLDocument
-from librarian import ParseError, DCNS
+from librarian import ParseError, DCNS, get_resource
 from librarian import functions
 
 
@@ -32,12 +32,12 @@ functions.reg_ends_white()
 functions.reg_texcommand()
 
 STYLESHEETS = {
-    'wl2tex': 'xslt/wl2tex.xslt',
+    'wl2tex': 'pdf/wl2tex.xslt',
 }
 
 
-def insert_tags(doc, split_re, tagname):
-    """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree 
+def insert_tags(doc, split_re, tagname, exclude=None):
+    """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree
 
     >>> t = etree.fromstring('<a><b>A-B-C</b>X-Y-Z</a>');
     >>> insert_tags(t, re.compile('-'), 'd');
@@ -46,6 +46,8 @@ def insert_tags(doc, split_re, tagname):
     """
 
     for elem in doc.iter(tag=etree.Element):
+        if exclude and elem.tag in exclude:
+            continue
         if elem.text:
             chunks = split_re.split(elem.text)
             while len(chunks) > 1:
@@ -65,15 +67,19 @@ def insert_tags(doc, split_re, tagname):
 
 
 def substitute_hyphens(doc):
-    insert_tags(doc, 
+    insert_tags(doc,
                 re.compile("(?<=[^-\s])-(?=[^-\s])"),
-                "dywiz")
+                "dywiz",
+                exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+                )
 
 
 def fix_hanging(doc):
-    insert_tags(doc, 
+    insert_tags(doc,
                 re.compile("(?<=\s\w)\s+"),
-                "nbsp")
+                "nbsp",
+                exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+                )
 
 
 def move_motifs_inside(doc):
@@ -122,24 +128,26 @@ def hack_motifs(doc):
 
 
 def parse_creator(doc):
-    """ find all dc:creator tags and add dc:creator_parsed with forenames first """
-    for creator in doc.findall('//'+DCNS('creator')):
-        p = Person.from_text(creator.text)
-        creator_parsed = deepcopy(creator)
-        creator_parsed.tag = DCNS('creator_parsed')
-        creator_parsed.text = ' '.join(p.first_names + (p.last_name,))
-        creator.getparent().insert(0, creator_parsed)
+    """ find all dc:creator and dc.contributor tags and add *_parsed versions with forenames first """
+    for person in doc.xpath("|".join('//dc:'+(tag) for tag in (
+                    'creator', 'contributor.translator', 'contributor.editor', 'contributor.technical_editor')),
+                    namespaces = {'dc': str(DCNS)})[::-1]:
+        if not person.text:
+            continue
+        p = Person.from_text(person.text)
+        person_parsed = deepcopy(person)
+        person_parsed.tag = person.tag + '_parsed'
+        person_parsed.set('sortkey', person.text)
+        person_parsed.text = p.readable()
+        person.getparent().insert(0, person_parsed)
 
 
-def get_resource(path):
-    return os.path.join(os.path.dirname(__file__), path)
-
 def get_stylesheet(name):
     return get_resource(STYLESHEETS[name])
 
 
 def package_available(package, args='', verbose=False):
-    """ check if a verion of a latex package accepting given args is available """  
+    """ check if a verion of a latex package accepting given args is available """
     tempdir = mkdtemp('-wl2pdf-test')
     fpath = os.path.join(tempdir, 'test.tex')
     f = open(fpath, 'w')
@@ -158,7 +166,7 @@ def package_available(package, args='', verbose=False):
     return p == 0
 
 
-def transform(provider, slug=None, file_path=None, 
+def transform(provider, slug=None, file_path=None,
               output_file=None, output_dir=None, make_dir=False, verbose=False, save_tex=None, morefloats=None):
     """ produces a PDF file with XeLaTeX
 
@@ -221,7 +229,7 @@ def transform(provider, slug=None, file_path=None,
 
         # LaTeX -> PDF
         shutil.copy(get_resource('pdf/wl.sty'), temp)
-        shutil.copy(get_resource('pdf/wl-logo.png'), temp)
+        shutil.copy(get_resource('res/wl-logo.png'), temp)
 
         cwd = os.getcwd()
         os.chdir(temp)
@@ -263,7 +271,7 @@ def transform(provider, slug=None, file_path=None,
 
 
 def load_including_children(provider, slug=None, uri=None, file_path=None):
-    """ makes one big xml file with children inserted at end 
+    """ makes one big xml file with children inserted at end
     either slug or uri must be provided
     """
 
@@ -276,14 +284,15 @@ def load_including_children(provider, slug=None, uri=None, file_path=None):
     else:
         raise ValueError('Neither slug, URI nor file path provided for a book.')
 
-    document = WLDocument.from_file(f, True,
-        parse_dublincore=True,
-        preserve_lines=False)
+    text = f.read().decode('utf-8')
+    text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
 
-    f.close()
+    document = WLDocument.from_string(text, True,
+        parse_dublincore=True)
 
+    f.close()
     for child_uri in document.book_info.parts:
+        print child_uri
         child = load_including_children(provider, uri=child_uri)
         document.edoc.getroot().append(child.edoc.getroot())
-
     return document